Fuel cells which are currently of commercial interest operate on streams of pure or nearly pure hydrogen, which is not readily available in most vehicles. Neither is a source of pure hydrogen convenient or safe to carry on board commercial trucks, buses or other vehicles. However, liquid hydrocarbons, such as diesel fuels, are easily available and their handling, storage and distribution are well developed. Consequently, the large-scale use of fuel cells is expected to require conversion of liquid fuels into a stream of pure hydrogen or hydrogen/CO2 mixtures, with only trace amounts of CO or sulfur impurities. This conversion will require a multi-step process to be carried out on board vehicles.
Dramatic progress has been observed in fuel-cell technologies in recent years. A prototype fuel cell-powered bus has been built by Ballard Power Systems for Vancouver's BC Transit. In this bus, compressed hydrogen is used to fuel the cells, which has raised concerns about passengers' safety. In a different venture, Argonne National Laboratory has built three prototype buses running on fuel cells. These vehicles operate with the diesel engine replaced by an electric engine, a phosphoric-acid fuel cell, and an on-board reformer. The role of the reformer is to convert liquid methanol into hydrogen in situ, and thus to avoid the necessity of carrying pressurized hydrogen. It is interesting to note that Argonne's fuel cell and the reformer are not much larger than the diesel engine they replaced. The fact that methanol is not currently a widely used fuel poses obvious limitations. There are also concerns related to long-term viability as well as corrosiveness and toxicity of methanol.
The development of an on-board system capable of converting hydrocarbon fuels, such as gasoline, diesel, JP-5, natural gas, etc., into a stream of hydrogen-rich gas would make it possible to power vehicles using standard fuels in combination with fuel cells. This would greatly accelerate the introduction of fuel-cell technologies into mass transit and help reduce air pollution in urban centers (particulates, NOx, CO, and unburned hydrocarbons). The advantage of on-board fuel processing is clear: the utilization of conventional fuels at improved efficiency, lower pollution levels, and zero noise.
Partial Oxidation—One current approach to the conversion of standard liquid fuels into hydrogen is partial oxidation (POX) of the liquids to produce soot, carbon oxides and hydrogen. The reaction is normally carried out without a catalyst in the temperature range 1100-1500° C. This technology is similar to the process used in the manufacture of carbon black (Austin, G. T., Shreve's Chemical Process Industries, Fifth edition, McGraw-Hill, New York, 1984). A number of projects are currently under way in which fuel processors based on partial oxidation are being developed (Preprints of the Annual Automotive Technology Development Contractors' Coordination Meeting, PNGV Workshop on Fuel Processing for Proton Exchange Membrane (PEM) Fuel Cells, Dearborn, Mich., 23-27 Oct., 1995, Office of Transportation Technologies, U.S. Department of Energy, Washington, D.C., 1995; Preprints of the Annual Automotive Technology Development Contractors' Coordination Meeting, vol. I, Dearborn, Mich., 23-27 Oct., 1995; “Recent Advances in Fuel Cells,” M. A. Wójtowicz, Symposium Organizer, in ACS Div. of Fuel Chemistry Prepr. 44 (4), pp. 972-997, 1999; “Hydrogen Production, Storage, and Utilization,” C. E. Gregoire-Padro and F. S. Lau, Symposium Organizers, in ACS Div. of Fuel Chemistry Prepr. 44 (4), pp. 841-971, 1999). The advantages of partial oxidation include simplicity, exothermicity of the process, sulfur tolerance, rapid start-up, rapid response to load changes, and compactness. However, partial oxidation produces relatively small amounts of gaseous hydrogen, which is diluted with nitrogen, large amounts of carbon oxides and soot, and the efficiency of fuel utilization is relatively low.
Steam Reforming—A second approach is based on steam reforming of hydrocarbon fuels according to the following reaction:
 CnHm+nH2O------>nCO+(n+m/2) H2  (A)
wherein n and m are typically in the range 1-20 and 4-42, respectively.
Since the above reaction is endothermic, the unreacted hydrogen from the fuel cell is usually burned to provide process heat. The reaction occurs over a catalyst in the temperature range 700-1000° C.
Since proton-exchange membrane (PEM) fuel cells, which are typically used in transportation applications, are intolerant to carbon monoxide, the latter species present in the product gas is often shifted to carbon dioxide according to the following reaction:CO+H2O<------>CO2+H2  (B)
Shift conversion is usually carried out in two stages: a high-temperature stage followed by a low-temperature stage. The former stage promotes high reaction rates, whereas the low-temperature stage increases the yield. Since the water-gas shift reaction is exothermic, inter-stage cooling is often implemented. In high-temperature fuel cells, CO can be oxidized to CO2 directly, and no shift reaction is necessary.
Steam reforming is a well-established large scale technology, but design, construction, and operation of compact reformers is quite a challenge. Common feedstocks for steam reforming are natural gas, propane and butane. The use of heavier feedstocks, such as naphtha, is difficult, and this problem can be only partly alleviated by the use of specially prepared catalysts (Austin, G. T., Shreve's Chemical Process Industries, Fifth edition, McGraw-Hill, New York, 1984). In most cases, a desulfurization step is required upstream of the reformer to protect catalyst beds from deactivation.
Autothermal Reforming—Autothermal reforming (ATR) is a hybrid approach involving endothermic steam reforming combined with partial oxidation for heat generation. The fuel is mixed with a mixture of steam and air, preheated, and fed into a catalytic reactor. Proper control of the steam-to-fuel ration is required to avoid coke formation, and the reaction usually occurs at 650-700° C. The effluent is typically sent to a shift reactor prior to entering the fuel cell. The advantages of ATR include compactness, and nitrogen dilution is the main disadvantage. The efficiency of ATR is lower than that of steam reforming but higher than POX.